Developmental Cognitive Neuroscience 60 (2023) 101198

Contents lists available at ScienceDirect

Developmental Cognitive Neuroscience

journal homepage: www.elsevier.com/locate/dcn

Characterizing different cognitive and neurobiological profiles in a

community sample of children using a non-parametric approach: An
fMRI study
Victoria Khalfin Fekson a, Tomer Michaeli b, Keri S. Rosch c, d, Bradley L. Schlaggar c, e,
Tzipi Horowitz-Kraus a, c, d, f, *
a
Educational Neuroimaging Group, Faculty of Education in Science and Technology, Technion, Israel
b
Faculty of Electrical and Computer Engineering, Technion, Israel
c
Kennedy Krieger Institute, Baltimore, MD, USA
d
Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, USA
e
Departments of Neurology and Pediatrics, Johns Hopkins University School of Medicine, Baltimore, MD, USA
f
Faculty of Biomedical Engineering, Technion, Israel

A R T I C L E I N F O A B S T R A C T

Keywords: Executive Functions (EF) is an umbrella term for a set of mental processes geared towards goal-directed behavior
Inhibitory control supporting academic skills such as reading abilities. One of the brain’s functional networks implicated in EF is
Default Mode Network the Default Mode Network (DMN). The current study uses measures of inhibitory control, a main sub-function of
Latent Profiles
EF, to create cognitive and neurobiological "inhibitory control profiles" and relate them to reading abilities in a
Nonparametric approach
Reading abilities
large sample (N = 5055) of adolescents aged 9–10 from the Adolescent Brain Cognitive Development (ABCD)
study. Using a Latent Profile Analysis (LPA) approach, data related to inhibitory control was divided into four
inhibition classes. For each class, functional connectivity within the DMN was calculated from resting-state data,
using a non-parametric algorithm for detecting group similarities. These inhibitory control profiles were then
related to reading abilities. The four inhibitory control groups showed significantly different reading abilities,
with neurobiologically different DMN segregation profiles for each class versus controls. The current study
demonstrates that a community sample of children is not entirely homogeneous and is composed of different
subgroups that can be differentiated both behaviorally/cognitively and neurobiologically, by focusing on
inhibitory control and the DMN. Educational implications relating these results to reading abilities are noted.

1. Introduction divided into cognitive inhibition (the ability to filter out unnecessary
stimuli to stay tuned on the task and focus attention) and behavioral
1.1. Cognitive development and inhibitory control across development inhibition (the ability to control emotions and impulsivity) (Diamond,
2013; Fox et al., 2005).
Executive Functions (EF) comprise a set of mental processes geared The age window of 9–10 years is an interesting one, both from the EF
toward goal-directed behavior and self-control, problem-solving, and and academic abilities (i.e. the reading) developmental perspectives. At
planning (Diamond, 2013). EF includes three main sub-functions: this developmental age window, children achieve similar levels of EF to
working memory, shifting, and inhibition (Diamond, 2013), with re­ adults in several EF, which can be clustered under response speed (speed
ports of inhibition abilities as related to academic abilities such as of processing), set maintenance, and planning (Welsh et al., 1991).
reading (Blair and Razza, 2007; Borella et al., 2010; Doyle et al., 2018; However, a fine-grained developmental study pointed at the develop­
Meixner et al., 2019), math (Clark et al., 2010; Cragg and Gilmore, 2014; mental trajectory of subcomponents of EF within a slightly extended age
Espy et al., 2004), and emotional regulation (Bartholomew et al., 2021). window (3–12 years old children), showing how inhibitory control de­
According to Diamond, inhibition is one of the central EF, as it is also velops first, followed by selective attention, and then more complex EFs

* Correspondence to: Educational Neuroimaging Group, Faculty of Education in Science and Technology, Faculty of Biomedical Engineering, Technion, Israel.
E-mail address: [email protected] (T. Horowitz-Kraus).

https://doi.org/10.1016/j.dcn.2023.101198
Received 9 December 2021; Received in revised form 6 December 2022; Accepted 11 January 2023
Available online 13 January 2023
1878-9293/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

(planning, fluency) develop into adolescence (Klenberg et al., 2001). As adulthood (63–73 years) decreased intra-network connectivity within
for reading achievement, according to Chall’s developmental model for the DMN was observed as well as decreased segregation with additional
milestones in reading development (Chall, 1983), by the age of 9 years, EF networks associated with declined cognitive abilities (Ng et al.,
children are transitioning from “learning to read” to “reading to learn”. 2016). This decreased intra-network FC with age was also related to the
This phase requires automatic reading skills so attention and cognitive speed of processing and episodic memory (Staffaroni et al., 2018), also
resources can be “released” for reading comprehension (see also supported by others in younger adults (Esposito et al., 2018).
(LaBerge and Samuels, 1974)). Interestingly, prior studies have suggested that different parts of the
DMN are potentially related unequally to a variety of developmental/
1.2. Inhibitory control and reading skills maturational changes (Fair et al., 2008; Fan et al., 2021; Supekar et al.,
2010), semantic processing differences (Seghier and Price, 2012),
Notably, although inhibitory control is one of the first EF to develop, attention challenges (Fan et al., 2018; Rohr et al., 2018), cognitive
reading and comprehension rely heavily on this cognitive ability. It was disorders (Sripada et al., 2014b; Swanson et al., 2011) and the func­
suggested that inhibition is an important cognitive skill contributing to tionality of the motor system (Hanakawa et al., 2003; Malouin et al.,
individual differences in cognitive, emotional, academic abilities and 2003; Margulies et al., 2009; Uddin et al., 2009; Zhang et al., 2014;
creativity (Harnishfeger and Bjorklund, 1994). Regarding academic Zhang and Li, 2012). In relation to the activation/FC patterns of the
outcomes, inhibition abilities play a role in several components underly DMN, an overall increased deactivation of this network was found in
reading; per the Simple View of the Reading model, language processing children ages 5–18 years old, while listening to stories, with more
and word decoding development lead to reading comprehension (Gough extensive deactivation in older children, associated with better task
and Tunmer, 1986). Inhibitory control was found to be related to comprehension scores (Horowitz-Kraus et al., 2017). More specifically
different reading components (word decoding (Spencer and Cutting, and concerning the range of the current study’s population, relations
2021; Taboada Barber et al., 2021); reading comprehension (Abo-elhija between the level of DMN and attention network (dorsal attention
et al., 2022; Butterfuss and Kendeou, 2018; Conners, 2009; Haft et al., network) and emotional and EF abilities in 9–10 years old children were
2019)), and although fluent reading is not officially part of the SVR found (Owens et al., 2020).
model (Adlof et al., 2006; Kim, 2020), this foundational reading skill An attempt to profile children based on their inhibition abilities has
was also associated with inhibition (Johann et al., 2020). not been conducted before but has been conducted using EF profiles
Due to its central role in EF and especially inhibition in academic using LPA. Dajani et al. (2016) used LPA for creating "EF profiles" using
outcomes (Harnishfeger and Bjorklund, 1994), the current study’s goal different cognitive tests to evaluate sub-levels of EF, such as working
was to distinguish groups of children based on inhibitory control and memory, inhibitory control, and cognitive flexibility, as indicators
their reading measures using computational tools (i.e. Latent Profile (Dajani et al., 2016). Dajani and colleagues divided a mixed group of
Analysis; LPA (Pastor et al., 2007; G. A. Williams and Kibowski, 2016; 8–13-year-old children identified as typically developing children,
Wurpts and Geiser, 2014). children with ADHD, or with Autism Spectrum Disorder (ASD), into
three sub-groups distinguished by their general EF levels (low, average,
1.3. The neurobiological profiles associated with inhibition skills and high). The result included three EF profiles, while each profile
included a mix of children from different sub-groups (Dajani et al.,
The neuroimaging literature converges on several functional net­ 2016). In a follow-up study conducted by these researchers, no differ­
works associated with EF, e.g. the Frontoparietal (Dosenbach et al., ences in FC between these EF sub-groups were observed (Dajani et al.,
2008; Ptak, 2012), Cingulo-opercular (Dosenbach et al., 2008; Neta 2019). The lack of findings in that study could be due to the use of
et al., 2014), Dorsal Attention, Ventral attention (Vossel et al., 2014)) parental reports only (BRIEF questionnaire) for creating the sub-profiles
and Salience networks (Seeley et al., 2007). Several papers focusing on or the use of several EF components instead of one sub-component. This
typically and atypically developing individuals (e.g., those with atten­ may cause more heterogeneous profiles and mask the common patterns.
tion deficit hyperactivity disorder or schizophrenia), suggest intrigu­ Additionally, the number of participants may not have been large
ingly that the Default Mode Network (DMN), is implicated in EF as it enough (N = 129, three groups of n = 43) to identify common functional
may also be important for inhibitory function (Fryer et al., 2018; patterns related to general EF ability strength.
Hernández-Álvarez et al., 2020; Liddle et al., 2011). The DMN is mainly Due to the involvement of inhibition in several emotional and
activated during rest in internally-directed cognitive processes such as cognitive deficits in childhood (Diamond, 2013; Fox et al., 2005), its
feeling processing, future planning, and retrieving memories (Buckner, central role in academic achievements (Blair and Razza, 2007; Borella
2012). The DMN is also involved in performing different experimental et al., 2010; Clark et al., 2010; Doyle et al., 2018; Espy et al., 2004) and
tasks, including social processing (e.g. such as emotional perception, especially in reading (Doyle et al., 2018) and reading comprehension
empathy, theory of mind, and morality) (Li et al., 2014), language (Borella et al., 2010), it is essential to detect individual differences in
(Horowitz-Kraus et al., 2017), semantic processing (Binder et al., 2005; inhibition abilities even among typically developing individuals.
Lanzoni et al., 2020; Seghier and Price, 2012; Wirth et al., 2011) and Hence, the goal of the current study is to determine the neurobio­
other EF processes such as attention (Leech et al., 2011; Rohr et al., logical correlates to different inhibitory control profiles in association
2018), planning (Spreng et al., 2010) and error monitoring (C. S. Li with reading abilities by focusing on the DMN and using a non-
et al., 2007). A handful of papers have outlined an altered involvement parametric approach in a community sample of children ages 9–10
of the DMN during rest in different clinical conditions such as Alz­ years (from the Adolescent Brain Cognitive Development (ABCD) data­
heimer’s disease (Buckner, 2012), Schizophrenia (Doucet et al., 2020; set). During this age range, proficient technical reading should be ach­
Fan et al., 2018), mood and anxiety disorders (Doucet et al., 2020) and ieved (Chall, 1983), and inhibition control is in its maturation period
Attention Deficit Hyperactivity Disorder (ADHD) (Sidlauskaite et al., (Klenberg et al., 2001), so differences between profiles will not be
2016; Sripada et al., 2014b). Although there is a considerable amount of attributed to psychiatric/developmental or neurological disorders. We
research on children with ADHD (Castellanos and Proal, 2012; Duffy hypothesized that different inhibition profiles would be revealed based
et al., 2021; Harikumar et al., 2021; Koziol et al., 2013), fewer studies on a variety of inhibition measures. We assumed divergent FC segre­
have focused on patterns of the DMN related to inhibitory control and gation patterns within the DMN would be found on each profile related
academic abilities such as reading, in a community sample of children, to what was previously reported (Fair et al., 2008; Fan et al., 2021; Fan
which is the topic of the current study. et al., 2018; Hanakawa et al., 2003; Rohr et al., 2018; Sripada et al.,
Moreover, developmental changes in the functional connections 2014a; Swanson et al., 2011; Uddin et al., 2009; Zhang et al., 2014). We
within the DMN were reported in different stages along the life span. In also hypothesized that profiles demonstrating lower inhibitory control

2
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

scores would also demonstrate lower reading abilities. Table 2

Summary of behavioral tests.
2. Methods Behavior Test Measures Mean Range
(SD)
2.1. Participants Inattention and Child Behavior DSM-5 ADHD t- 52.64 50–80
Hyperactivity Checklist score (5.09)
5055 children at ages 9–10 years (Mean (in months) = 119.83, SD = Cognitive Stop Signal Incorrect go trials – 0.079 0 – 0.90
Inhibition Task a proportion (0.073)
7.45; 2626 females) were selected from a cohort of 11,880 in the curated
Stop Signal Incorrect stop trials 0.44 0.03 –
annual release 2.0.1 of The Adolescent Brain Cognitive Development Task – a proportion (0.10) 0.98
(ABCD) study (https://abcdstudy.org/). The 11,880 participants Stop Signal Incorrect stop trials 164.5 32 –
recruited across 22 sites in the U.S (2018–2019) reflect the U.S popu­ Task - mean response (31.9) 285.7
lation of boys and girls with diverse races and ethnicities, education, time (ms)
Behavioral Behavioral Summary raw score 9.48 0–21
income levels, and living environments. Inclusion criteria for the ABCD
Inhibition Inhibition (3.62)
study were 9–10 years old children that are fluent in English, approved System
for MRI scanning, without a major neurological disorder, were born Impulsivity Behavioral Drive raw score 3.92 0–12
after 28 weeks, without a history of brain injuries and a diagnosis of Activation (2.96)
System
schizophrenia, moderate and above ASD, mental retardation or alcohol/
Behavioral Fun-seeking raw 5.60 0–12
substance use disorder (Karcher et al., 2019). Activation score (2.57)
From the entire sample of 11,880 participants, the participants System
included in the current study met the following criteria: 1) at least 10 Behavioral Reward 10.88 0–15
min usable resting-state scan; 2) no clinical diagnosis; 3) the stop-signal Activation responsiveness raw (2.88)
System score
task measures (Logan, 1994); 4) t-score for CBCL ADHD DSM5 Scale
Reading Oral Reading Total score, age- 103.99 68–206
(Achenbach and Ruffle, 2000); and 5) the BIS-BAS questionnaire raw Recognition corrected (18.78)
scores (Carver and White, 1994). The demographics and behavioral test Task
statistics can be found in Tables 1 and 2, respectively. The project and
hypothesis were not pre-registered.
relative homogenous scores across participants. Each test was related to
a different aspect of inhibitory control ability, as can be shown in
2.2. Behavioral data
Table 2. Several variables were derived from each test, resulting in total
eight indicators for the LPA algorithm.
2.2.1. LPA Indicators
Several tasks were used for profiling the participants in the current
2.2.1.1. Stop signal task. This task was held as part of fMRI task-based
study: Stop Signal Task (SST) (Logan, 1994), DSM-5 ADHD from the
scans (Casey et al., 2018). In this task, the participants were required
Child Behavior Checklist (CBCL) (Achenbach, 2009), and BIS-BAS
to press the right or the left button according to the arrow on the screen.
questionnaire (Carver and White, 1994). The Flanker test (Eriksen and
If a "Stop" stimulus was presented, they needed to withhold their reac­
Eriksen, 1974) was also initially selected but removed due to a lack of
tion and not press the button (Casey et al., 2018). They were instructed
contribution to profiling separation with the LPA algorithm due to
to react as quickly and accurately as possible. Incorrect go trial is defined
as a late response, pressing an incorrect button (i.e., left instead of right),
Table 1
or no response. An incorrect stop trial is defined as responding to a stop
Participant’s demographic summary.
trial by pressing the left or right button.
Variable Frequency Percentage The following task-generated variables were included in this
Sex research: the proportion of incorrect go trials, the proportion of incor­
Female 2626 51.95% rect stop trials, and the mean response time for all incorrect "Stop" trials.
Grade, age in months (mean, SD) The incorrect trial scores were selected so they would match in direc­
2nd grade (110.92, 2.57) 27 0.53%
3rd grade (110.81, 3.58) 750 14.8%
tionality with the other measures in this study.
4th grade (116.75, 5.39) 2189 43.3%
5th grade (126.11, 3.97) 1906 37.7% 2.2.1.2. Child’s ADHD symptom scale. The parents were asked to report
6th grade (129.56, 1.92) 182 3.6% the child’s behavior using the CBCL form (Achenbach, 2009). The
7th grade (129.00, 0) 1 0.19%
Handedness
DSM-5 ADHD t-scores were chosen from the CBCL, which are normal­
Right 4099 81.08% ized to age and gender.
Mixed 603 11.92%
Left 352 6.98% 2.2.1.3. Behavioral inhibition and behavioral activation. The children
Ethnicity/Race
White/Non-Hispanic 1972 58.79%
filled out the BIS-BAS questionnaire (Carver and White, 1994), and a
Hispanic 915 18.10% total of 4 raw scores variables were derived from this questionnaire:
Black 559 11.05% three aspects of the BAS system (drive, fun-seeking, and reward
Asian 89 1.79% responsiveness) and one for the BIS system.
Other 513 10.14%
Additional information regarding these inhibition tasks is listed in
Family Income (in US dollars)
> = 100 K 2149 42.51% Supplementary material 1.
50k < 100 K 1413 27.95%
< 50 K 1125 22.25% 2.2.2. Reading analysis
Didn’t know\refuse to answer 368 7.29% To determine differences in reading abilities between the sub-
Parental Education
Post graduate degree 1893 37.44%
profiles, a single-word reading recognition measure, the Oral Reading
Bachelor’s degree 1406 27.81% Recognition Task (Gershon et al., 2013; Luciana et al., 2018), from the
Some college 1245 24.62% NIH toolbox (NIH Toolbox (healthmeasures.net)) was used. In this task,
High school diploma or GED 360 7.12% participants were required to recognize words or letters presented on the
Didn’t finish high school 148 2.92%
screen. Each response was scored as correct or incorrect by the

3
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

examiner. A total score normalized by age (Casaletto et al., 2015) was acquisitions and adult-size coils acquired in 22 sites across the U.S (for
derived from this test. the sites see https://abcdstudy.org/contact/). The scanning parameters
for the fMRI images and structural images are summarized in Table S1 in
2.3. Neuroimaging data the Supplemental materials. Due to an error in processing Philips scans
in ABCD 2.0.1 collection (ABCD, 2019), these scans were omitted from
Resting-state fMRI derivatives from the ABCD-BIDS dataset (https:// this study.
collection3165.readthedocs.io/en/stable/derivatives/) were used to
determine the DMN in the current study. In the resting state task, the 2.5. Data analyses
participants were required to relax and stare at a fixation cross for a total
time of 20 min (in sets of 5 min each). Before scanning, they trained for 2.5.1. The latent profile analysis (LPA)
motion compliance in a mock scanner. During the scan, a real-time The latent profile analysis (LPA) was run using the tidyLPA library in
motion correction for structural scanning was applied. During the R version 4.0.2 software (Rosenberg et al., 2018) and rstatix library
resting-state scan, another real-time motion monitoring was applied, (Kassambara, 2020), using eight indicators representing inhibitory
called FIRMM (fMRI Integrated Real-Time Motion Monitor (Dosenbach control in different aspects (Table 2). Different models with a different
et al., 2017) allows a live adjustment of the scanning paradigm (Casey number of classes to identify the most suitable (1–6 latent classes) were
et al., 2018). tested. Each model was with equal variance and zero covariance. The
model was selected based on the following criteria (Dajani et al., 2016;
2.4. Data acquisition and scan parameters Nylund and Muthén, 2007; Tein et al., 2013): Bayesian Information
Criterion (BIC), Akaike Information Criterion (AIC), sample-adjusted
The ABCD neuroimaging data were acquired from three types of 3 T BIC (SABIC), Entropy and Likelihood Ratio Test (BLRT). In all the in­
scanners: Siemens Prisma, General Electric (GE) 750, and Philips formation criteria (BIC, AIC, and SABIC), lower values indicate better
(Achieva and Ingenia) with multiband echo-planar imaging (EPI) model fit. In the Entropy index, higher values indicate better class

Fig. 1. Creation of divisions. Fig. 1: In each version, the entire cohort was divided into two groups that matched the following variables: class distribution, age, sex,
race/ethnicity, handedness, parental education, and family income. In each division, a common correlation matrix was created for each of the classes and all subjects
in this division (n = 2527 or 2528).

4
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

separation. The BLRT, a parametric bootstrap method, provides a 2.5.2. Covariate analysis
p-value that can be used to compare the increase in model fit between Socio-Economics Status (SES), and motion (average Framewise
the k-1 and k-class model and the ratio from the sample in the smallest Displacement (FD)) variables were included in the LPA model and
class. To help the model selection decision, a graph of the BIC, AIC, and analyzed using the Mplus program version 8 (Muthén and Muthén,
SABIC values was added (see Fig. 3). The gradient in the graph repre­ 1998–, 2017). Differences in age and grade between the classes were
sented when the difference between the models was very small. After tested with appropriate statistical analysis and, due to insignificant
choosing the number of classes, the non-parametric Kruskal-Wallis test findings, were excluded from the covariate analysis. Table 3 represents
was run for each of the indicators to characterize the difference between the descriptive statistics of demographic variables, together with mo­
the classes. tion, between the classes.

Fig. 2. An example of the effect size calcula­

tion. Fig. 2: (a) An example for effect size
calculation for only one division. One effect size
table was created with only positive correla­
tions in both class1 and class2. ROI = Region Of
Interest; Zxy = correlation value after fisher Z
transformation between ROIx and ROIy. (b) An
example of effect size calculation is the output
for all divisions together. The effect size tables
from (a) were calculated for each division and
then merged. The output is a table of ten effect
size values for each comparison. esx = effect
size of x division.

5
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Table 3
Descriptive statistics of demographic and motion variables among classes.
Variable Class1 – Class2 – Class3 – Class4 –
Behavioral Controls Cognitive Inattention\
Inhibition n = 2661 Inhibition Hyperactivity
n = 1716 n = 332 n = 346

Age 119.78 (7.46) 119.89 119.94 119.49 (7.42)

(7.47) (7.30)
Motion 0.159 (0.105) 0.148 0.171 0.161 (0.098)
(average FD) (0.093) (0.099)
Sex
Female 861 (50.17%) 1476 127 162 (46.82%)
(55.46%) (38.25%)
Fig. 3. Graph of the information criteria of the LPA model. Fig. 3: X-axis rep­
Grade, age in
resents the number of classes. Y-axis represents the score of the information months
criteria. The blue color stands for BIC = Bayesian Information Criterion. Orange (mean, SD)
color stands for AIC = Akaike Information Criterion. Grey color stands for 2nd grade 11 (0.64%) 11 1 (0.3%) 4 (1.16%)
SABIC = sample-adjusted BIC. (110.92, (0.41%)
2.57)
3rd grade 262 (15.27%) 377 60 (18.07%) 51 (14.74%)
2.5.3. Association of the different profiles with reading decoding scores (110.81, (14.17%)
After creating the profiles, the Kruskal-Wallis test was conducted to 3.58)
explore the reading decoding scores (Gershon et al., 2013; Luciana et al., 4th grade 735 (42.83%) 1164 134 156 (45.09%)
2018) effect between the classes. The post-hoc Dunn test and a Bonfer­ (116.75, (43.74%) (40.36%)
5.39)
roni correction were conducted (p < 0.05/6 =p < 0.008).
5th grade 657 (38.29%) 1003 124 122 (35.26%)
(126.11, (37.69%) (37.35%)
2.5.4. Neuroimaging data processing 3.97)
The derivative neuroimaging data went through preprocessing steps 6th grade 51 (2.97%) 105 13 (3.92%) 13 (3.76%)
previously detailed in Hagler Jr et al. (2019). Average time courses of (129.56, (3.95%)
1.92)
333 Regions of Interest (ROI) were created according to Gordon atlas 7th grade 0 1 (0.04%) 0 0
(Gordon et al., 2016). Following the parcellation of the data, the current (129.00, 0)
study employed the 333 cortical ROIs from Gordon’s atlas and then Handedness
focused the analysis only on DMN per our study aims. To deal with Right 1393 (81.18%) 2163 270 273 (78.90%)
(81.29%) (81.33%)
excessive motion, Framewise Displacement (FD) was calculated (Power
Mixed 215 (12.53%) 300 44 (13.25%) 44 (12.72%)
et al., 2014), then frames with FD of more than 0.2 were excluded. We (11.27%)
used the Gordon atlas parcellation time series of the best 10 min scan Left 108 (6.29%) 198 18 (5.42%) 29 (8.38%)
frames, i.e. frames with the lower FD (see documentation: Collection (7.44%)
3165 - ABCD-BIDS Community Collection (ABCC)). Ethnicity/Race
White/Non- 905 (52.80%) 1687 206 174 (50.58%)
Hispanic (63.44%) (62.24%)
2.6. Accounting for subject-specific noise factors in functional Hispanic 371 (21.65%) 426 24 (16.31%) 45 (18.60%)
connectivity analysis (16.02%)
Black 241 (14.06%) 225 36 (10.88%) 57 (16.57%)
(8.46%)
The current study used a non-parametric approach which has fewer
Asian 36 (2.10%) 48 1 (0.3%) 4 (1.16%)
prior assumptions about a given population compared to parametric (1.81%)
methods (Corder and Foreman, 2009, p. 1–2). This non-parametric Other 161 (9.39%) 273 34 (10.27%) 45 (13.08%)
approach assumes no prior knowledge of the distribution of the data (10.27%)
and can be used to determine common patterns in functional MRI data Family Income
(in U.S
(Zhitnikov et al., 2018). Specifically, assuming that the covariance dollars)
matrix of the j-the subject is given by > = 100 K 618 (39.64%) 1298 140 93 (29.25%)
(51.88%) (45.45%)
Σxj = Σu + Σvj (1) 50k < 100 K 489 (31.37%) 724 100 100 (31.45%)
(28.94%) (32.47%)
where Σu is a component common to all subjects and Σvj is a subject- < 50 K 452 (28.99%) 480 68 (22.08%) 125 (39.31%)
specific component, the method provides an estimate for Σu . (19.18%)
Parental
Five correlation matrices were created for all 5055 subjects (one
Education
representing all subjects and one for each class derived from the LPA Post graduate 570 (33.26%) 1112 124 87 (25.14%)
analysis, see Results), using the Zhitnikov et al. algorithm (Zhitnikov degree (41.80%) (37.35%)
et al., 2018). Second, to reinforce reproducibility as recently encouraged Bachelor’s 472 (27.54%) 753 98 (29.52%) 83 (23.99%)
(Botvinik-Nezer et al., 2020; Klapwijk et al., 2021; Marek et al., 2020), degree (28.31%)
Some college 470 (27.42%) 571 77 (23.19%) 127 (36.71%)
the full cohort was randomly divided into two groups. Then, the two (21.47%)
groups were matched by the following descriptive criteria: class distri­ High school 137 (7.99%) 160 29 (8.73%) 34 (9.83%)
bution (following the LPA analysis), age, gender, race, handedness, diploma or (6.02%)
parent’s education, and household income)similar to the ABCD repro­ GED
Didn’t finish 65 (3.79%) 64 4 (1.20%) 15 (4.34%)
ducible matched samples (ARMS) (Feczko et al., 2020)). This process
high school (2.41%)
was performed five times and at the end, ten divisions were created
(Fig. 1). In each division, the Zhitnikov algorithm was run on the Note. FD = Framewise Displacement.
time-courses data. For each subject, a covariance matrix of the 333 ROIs
was calculated using the Ledoit-Wolf estimator (Ledoit and Wolf, 1996).

6
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

2.6.1. Calculating the differences in DMN between classes significantly contributed to the class separation with a least moderate
A comparison between each class to the group characterized by the magnitude. See Fig. 4 for the class separation.
higher inhibition control abilities (class2) referred to as the "control" in
the current study (e.g. class1-class2, class3-class2, class4-class2) was 3.1.2. Covariate analysis results
conducted in each division. The effect size was calculated, z1 – z2 Covariate analysis results are presented in Table 5 when each class is
(Cohen, 1977), of the difference in each ROI (see example in Fig. 2a). tested against class2 – controls.
This process was repeated for each division, with an overall of ten effect Based on this analysis, some covariates produced significant differ­
size values for each ROI in each comparison (e.g. class1-class2, ences across profiles, including household income and motion level.
class3-class2, class4-class2) (Fig. 2b). After conducting a significance Results suggest statistically significant differences between class2
test for comparing two correlation coefficients (Diedenhofen and Musch, (controls) and both class1 (behavioral inhibition) and class4 (cognitive
2015; Lenhard and Lenhard, 2014), ROIs with effect sizes smaller than inhibition) in household income Fig. 5 demonstrates the differences, in
0.1 were marked as significant, which is ordinary for a large sample size class2 more subjects have a household income in the upper level (>
(Khalilzadeh and Tasci, 2017; Marek et al., 2020). To give importance to 100 K) than in class1 and class4. Moreover, more subjects in class1 and
replicable results, only results that repeated at least in 90% of the di­ class4, had a household income at the lowest level (< 50 K annually).
visions were chosen. Thresholds of 0.1, 0.2, and 0.3 were tested, but only Motion, as measured using FD, also showed a statistically significant
with a threshold of 0.1, a variety of replicable results were demon­ difference between class2 (controls) and all the other classes.
strated. Moreover, negative correlations were also included in the
analysis, but the results were not replicable among the subdivisions; 3.1.3. Reading ability effect between the classes
hence, only the positive correlations are presented in this study. In Statistically significant differences in reading scores between the
conclusion, ROIs with an effect size greater than 0.1 repeated at least in four classes were found [F(3) = 55.719, η2 = 0.0105, p < 0.001]. The
90% of the divisions were selected. highest reading scores were found for the control group (class2),
whereas the lowest scores were found for the inattentive/hyperactive
3. Results group (class4). Post-hoc Dunn tests using a Bonferroni adjustment were
conducted to compare all pairs of classes. The difference between the
3.1. Behavioral results control class (class2) and the behavioral inhibition class (class1) was
significant. The inattention\hyperactivity class (class4) was different
3.1.1. Latent profile analysis results from all other classes. Classes 1 and 3 and classes 2 and 3 did not show
According to the BLRT significance parameter, all six models tested significant differences in their reading abilities. The means, standard
were significant (Table 4). The information criterion AIC/BIS/SABIC scores (including age-normalized reading scores), and post-hoc test re­
values decreased as the number of classes increased. The gradients from sults are described in Fig. 6 (For the full scores, see Table S4 and S5).
models 4–5 and from models 5–6, were the smallest (Fig. 3), and hence
models 4 and 5 were marked as the best choices. The entropy values of 3.2. Neuroimaging results
models 4 and 5 were 0.815 and 0.840, respectively, indicating a better
class separation for model 5. The sample ratio in the smallest class in 3.2.1. Comparison of the DMN ROIs in each class vs controls
models 4 and 5 were 0.0657 (332 subjects) and 0.0059 (29 subjects), To further investigate the characteristics that uniquely differentiate
respectively; hence model 4, four classes, was chosen. each class from the others, each class was compared to the control class,
The first class (N = 1716, 33.95%) had overall higher scores in the as described in the Methods section. Then, ROIs with effect size values
BIS-BAS questions, which represents behavioral activation (or disinhi­ that were greater than 0.1 and repeated in at least 90% of the divisions
bition) in response to rewards and behavioral inhibition in response to were selected. See Fig. 7 for the common correlation matrices composed
punishment (Carver and White, 1994). In general, this group is char­ of all Gordon networks and Fig. 8 for the DMN’s sub-matrices. In class1,
acterized by difficulties in behavioral inhibition and hence was cate­ which was associated with lower behavioral inhibition ability, greater
gorized as the Behavioral Inhibition class. The second class (N = 2661, FC was found between the anterior and the posterior parts of the brain in
52.64%) had lower scores in all measures (i.e. higher abilities in all the left hemisphere (Fig. 9a, Table S6, and Table S7). In class3, which
measures tested) and hence was categorized as the Control class. The was associated with lower cognitive inhibition ability, lower FC was
third class (N = 332, 6.57%) had high scores in the SST measures, rep­ found in the anterior part of the brain in both the left and right hemi­
resenting poorer cognitive inhibitory control (Logan, 1994) and hence spheres (Fig. 9b, Table S8, and Table S9). In class4, which was associated
was categorized as the Cognitive Inhibition class. The fourth class with lower attention, hyperactivity, and lower inhibition abilities
(N = 346, 6.84%) had the highest scores in inattention and hyperac­ (ADHD characteristics), greater FC was detected mostly with the pre­
tivity (measured using the CBCL ADHD scale) and relatively higher cuneus: left dorsal precuneus (parcel 94, Brodmann area 7), left ventral
scores in the other measures (cognitive and behavioral inhibition) and precuneus (parcel 1, Brodmann area 23, 31) and right ventral precuneus
hence was categorized as the Inattention\Hyperactivity class (ADHD (parcel 162, Brodmann area 23,31) with other regions within the DMN:
patterns). The variables’ means for each class are listed in Table S2 in left mPFC, right Middle temporal gyrus, right Cingulate Gyrus and right
Supplement materials. The results from the Kruskal-Wallis, shown in Superior Frontal Gyrus (Fig. 9c, Table S10, and Table S11). Greater
Table S3 in Supplement materials, demonstrate that each variable connectivity was also shown within the right hemisphere (Fig. 9c,

Table 4
Latent Profile Analysis results for models’ comparison.
Classes AIC BIC SABIC Entropy n_min BLRT_val BLRT_p

1 114788 114892 114841 1 1 NA NA

2 111794 111957 111878 0.674 0.399 3011 0.0099
3 109019 109241 109133 0.776 0.076 2728 0.0099
4 106677 106958 106821 0.815 0.065 2429 0.0099
5 105505 105845 105679 0.840 0.005 1773 0.0099
6 105424 105822 105628 0.720 0.039 442 0.0099

Note. AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion; SABIC = sample-adjusted BIC; n_min = ratio from the sample in the smallest class;
BLPT_val = value of the Bootstrap Likelihood Ratio Test; BLRT_P = significant of the Bootstrap Likelihood Ratio Test.

7
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Fig. 4. LPA four classes separation. Fig. 4: (a) This graph represents model 4 class separation. The X-axis represents the measures. The Y-axis represents the mean of
each measure, normalized. The red color stands for class1 (self-reported motivated inhibition), the green line stands for class2 (which we referred to as “controls”),
the light blue stands for class3 (task-based cognitive inhibition), and the purple color stands for class4 (parent-rated problems with inattention\hyperactivity).
BAS_DRIVE = BAS (behavioral activation system) score for Drive scale (pursing after the desired goal); BAS_FS = BAS (behavioral activation system) score for Fun-
seeking scale (desire for new rewards and spontaneity; BAS_RR = BAS (behavioral activation system) score for Reward responsiveness (positive reactions to an event
or expectation of reward); BIS_SUM = BIS (behavioral inhibition system) score; CBCL_ADHD = DSM-5 ADHD t-scores from the CBCL questionnaire; SST_INCRGO_RT
= rate of GO trials that were answered incorrectly in the Stop Signal Task; SST_INCRS_MRT = incorrect STOP trials mean response time (in ms) in the Stop Signal
Task; SST_INCRS_RT = rate of STOP trials that were answered incorrectly in the Stop Signal Task. (b) Classes distribution. X-axis represents the classes. Y-axis
represents the number of subjects in each class. The red color stands for class1 (behavioral inhibition), the green line stands for class2 (controls), the light blue stands
for class3 (cognitive inhibition), and the purple color stands for class4 (inattention\hyperactivity).

processing (Binder et al., 2005; Lanzoni et al., 2020; Seghier and Price,
Table 5
2012; Wirth et al., 2011). The results supported our hypotheses showing
Covariate analysis results.
different sub-profiles of inhibitory control and related neurobiological
Variable Class1 – Class3 – Class4 – Inattention/ differences in DMN. More specifically, high FC between anterior and
Behavioral Cognitive Hyperactivity
posterior DMN is related to the behavioral inhibition class, low FC
Inhibition Inhibition
within anterior DMN is related to the cognitive inhibition class, and
Parent -0.113 -0.076 -0.174
higher FC of the precuneus together with high FC in the right hemi­
education
Household -0.345 * ** -0.103 -0.580 * ** sphere related to the inattention\hyperactivity class. As postulated,
income different reading abilities were also found for the four profiles: the
Motion (FD) 1.143 * * 2.201 * ** 1.554 * * control class showed the highest reading scores, whereas the inattenti­
Note. Class 2 (Controls) served as a reference class. FD = Framewise Displace­ ve/hyperactive class showed the lowest scores.
ment.
* **p < 0.001, * * p < 0.001, * p < 0.01
4.1. Neurobiological differences between cognitive profiles
Table S10, and Table S11).
Our results demonstrate that neurobiological profiles can be identi­
fied for groups that were created solely on behavioral measurements.
4. Discussion
These results, though encouraging, are different from previous findings
(Dajani et al., 2019), although there are several main differences be­
The goal of the current study was to demonstrate, in a community
tween the studies. It may be that categorical differences between EF
sample of children, the relationship between heterogeneity in inhibitory
abilities ("Low", "Average", and "Above Average") reported in (Dajani
control, reading abilities, and connectivity features of a brain system
et al., 2019) are based solely on parental reports (rather than on per­
linked to inhibition and executive function. We focused on behavioral
formance tasks). In the current study, in addition to a parental ques­
measures associated with inhibition and the DMN, a functional network
tionnaire (CBCL ADHD DSM5), a child questionnaire (BIS-BAS
implicated in EF and inhibitory control (Fryer et al., 2018; Hernán­
questionnaire) and performance task (SST) were used as well, providing
dez-Álvarez et al., 2020; Liddle et al., 2011) and linguistic/semantic
a multi-informant and multi-method assessment of inhibitory control.

8
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Fig. 5. Distribution of the household income variable among classes. Fig. 5: The X-axis represents each class. The Y-axis represents the percentage of each of the
household income values.

Fig. 6. Reading ability per class and differences between classes. Fig. 6: X-axis represents each class. Y-axis represents the mean number of correct words, normalized
by age. The post-hoc Dunn tests significant results are demonstrated above the bars. * ** p < 0.0001, * * p < 0.001, * p < 0.01.

Also, while Dajani’s study used the BRIEF general EF questionnaire, here inhibition abilities, it can reflect an immature brain development.
we used only the inhibition EF sub-component, which likely created Studies investigating the development of DMN found that overall DMN
more homogenous profiles. Lastly, the current study used a novel FC increases with age (Fair et al., 2008; F. Fan et al., 2021) and
non-parametric analysis (Zhitnikov et al., 2018) aiming to reveal com­ particularly in anterior regions (F. Fan et al., 2021). Cognitive inhibition
mon phenomena that are masked in traditional analyses such as ICA. ability improves with age (Harnishfeger, 1995) and its improvement is
The benefits of this analysis (Zhitnikov et al., 2018) need to be tested connected with frontal lobe development (Harnishfeger, 1995). Hence,
further on different datasets to strengthen this claim. one of the reasons for delayed brain maturation may derive from frontal
lobe FC which may explain the low performance in this group.
Other researchers found that lower FC in anterior DMN is associated
4.2. Maturation of the DMN and inhibition profiles: early maturation is with attention difficulties in different attention domains (executive
related to a greater behavioral inhibition in response to punishment and attention and sustained attention) (J. Fan et al., 2018; Rohr et al., 2018).
reward abilities Attention and inhibitory control are related to each other, as to maintain
attention and to direct attention to the relevant stimulus, the brain needs
Our results suggest that children with greater behavioral inhibition to inhibit attention towards unnecessary external distractors (Diamond,
in response to punishment and a greater behavioral approach in reward 2013). Hence, we suggest that children in this class may also be cate­
than controls demonstrate higher FC between anterior and posterior gorized with low attention abilities and a future study including atten­
DMN. Previous studies suggested those connections as a signature of tion measures needs to be conducted in this class to strengthen the
brain maturation (Fair et al., 2008; F. Fan et al., 2021; Supekar et al., connection between inhibitory control and attention-related neurobio­
2010). Brain maturation can suggest a reduced brain plasticity, i.e. a logical profiles.
lower ability to adjust to different environments (Kolb and Gibb, 2011),
as brain plasticity is known to decline with age (Kolb and Gibb, 2011; Lu
et al., 2004). 4.3. Right lateralized functional connectivity of the DMN and attention
Amongst the four groups, the group of children with lower cognitive challenges
inhibition ability demonstrated a lower FC in the anterior DMN
compared to controls. In contrast to the group with lower behavioral The group with increased inattention\hyperactivity demonstrated

9
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Fig. 7. Common correlation matrices for all brain networks (based on the Gordon Atlas). Fig. 7: (a) Correlation matrix of all 5055 subjects. (b) Separate correlation
matrix for each class.

10
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Fig. 8. Common correlation matrices of the default mode network. Fig. 8: (a) Correlation matrix of all 5055 subjects. (b) Separate correlation matrix for each class.

stronger FC in the right hemisphere compared to controls (hyper­ et al., 2013), and connects lower attention abilities with right laterali­
connectivity of the DMN in ADHD was also recently found in (Duffy zation of the DMN (Sripada et al., 2014b). In general, the right side of the
et al., 2021)). These results are in line with previous research claiming brain was previously found to be associated with attention and inhibi­
that the DMN is mostly left-lateralized (Agcaoglu et al., 2015; Nielsen tion (Aron, 2007; Aron et al., 2004; Corbetta et al., 2008; Corbetta and

11
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Fig. 9. DMN differences in each class with the control class. Fig. 9: Brain images representing the correlations in each class that differ from the controls class with an
effect size greater than 0.1. These results were repeated in at least 90% of the divisions. The blue line represents a connection in which the correlation value in the
controls class was higher than the value in class X (X = 1,3,4). The red line represents a connection in which the correlation value in class X (X = 1,3,4) was higher
than the value in the control class.

Shulman, 2002; Singh-Curry and Husain, 2009). The right Inferior Pa­ study involving the DMN along with other brain regions related to
rietal Lobule (IPL), which is part of the DMN, is associated with both attention and inhibition can be conducted to determine the connection
maintaining attention and reorienting attention to new stimuli (Corbetta in this class with attention and inhibition control symptoms.
et al., 2008; Husain and Nachev, 2007; Singh-Curry and Husain, 2009).
Another network that is related to reorienting attention is the right 4.4. Higher involvement of the precuneus is related to the motor system,
ventral Frontoparietal network (Corbetta and Shulman, 2002). The right
and its deactivation is related to visual attention abilities
Inferior Frontal Gyrus (IFG), on the other hand, was related to inhibition
control (Aron, 2007; Aron et al., 2004). Hence, we suggest that higher
This inattention/hyperactivity class demonstrated a greater FC be­
FC in the right hemisphere is due to compensation with deficits in tween the left dorsal, left ventral, and right ventral Precuenues (parcel
attention and inhibitory control, which characterized this class. A future
94, parcel 1, and parcel 162, respectively), and brain regions associated

12
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

with EF and language processing (e.g. left mPFC, right Middle Temporal ability was found to vary depending on different parameters that may
Gyrus, right Cingulate Gyrus, and right Superior Frontal Gyrus). The not be evaluated when traditionally examining SES (i.e. children from a
ventral part of the precuneus was previously reported as negatively low SES background with a single parent showed lower inhibition
correlated with the motor network (Uddin et al., 2009; Zhang and Li, abilities than children from a comparable low SES background who have
2012), in contrast to the dorsal precuneus, which was positively corre­ two parents(Sarsour et al., 2011)). In addition, variable inhibition
lated with the motor system (Margulies et al., 2009; Zhang et al., 2014; abilities (and developmental trajectories in inhibition abilities) were
Zhang and Li, 2012). In the context of this class, it may be that greater found in children from a low socio-economic background(Pacheco et al.,
involvement of this part of the DMN may be related to an altered ability 2018). It is, therefore, possible that the fact that the demographic
to control the motor network as part of the higher hyperactive symptoms background of class2 is high is incidental. Hence, to determine if high
characterizing this group. inhibition abilities and specific profiles of DMN are truly characterized
In addition to the FC pattern related to the motor system, the dorsal by higher SES, a future study examining the differences in neurobio­
part of the precuneus was found to be triggered by motor-imaginary logical signature for DMN in children matched for their inhibition pro­
tasks (Hanakawa et al., 2003; Malouin et al., 2003). Lower perfor­ files from high vs low socio-economic status should be conducted.
mance in these tasks was found to be related to ADHD symptomatology
(Williams et al., 2013). Hence, a greater involvement of the dorsal part 4.7. Reading abilities are related to patterns of inhibitory control
of the precuneus may be used as a compensating mechanism for this
lower performance mentioned above. This is in line with findings Our results provide a strong link between inhibitory control and
showing that greater deactivation of the posterior precuneus during a reading abilities in early adolescents. These findings, by themselves, are
perceptual matching task compared to speech production is related to an not surprising and echo previous literature demonstrating the link be­
engagement of visual attention (Seghier and Price, 2012). The authors tween reading abilities (Allan et al., 2014) and reading comprehension
also state that the relation between age (maturation) and the level of and inhibitory control (Borella et al., 2010). The link to alterations of
deactivation of this region is affected by the level of visual attention and different parts of the DMN is also not surprising in light of the studies
perceptual processing demands of the task. This may be a possible supporting its role in semantic processing (Binder et al., 2005; Lanzoni
explanation also for the lower reading ability observed in this group of et al., 2020; Seghier and Price, 2012; Wirth et al., 2011). However, the
children in the current study. results show different reading abilities across the sub-profiles found in
the latent profile analyses. The control group showed the highest
4.5. What is the inhibitory profile of children characterized by ADHD reading scores, and the inattentive/hyperactivity group showed the
patterns? lowest. Also, significant differences were found between the controls
and the behavioral inhibition group (class1). A possible explanation for
Note that class4 is characterized by ADHD patterns (i.e., parent-rated these lower reading scores may derive from the neuroimaging data and
problems with attention, hyperactivity, and lower inhibitory control are related to the FC changes found in left-hemispheric regions within
abilities), but the children in this class were not recruited as a clinical the DMN in classes 1 and 4 compared to the control group. Previous
sample with a diagnosis of ADHD. The t-scores of the CBCL ADHD DSM5 studies did suggest a connection between the recruitment of regions
in this class are varied, as 104 (30%) are in the normal range (64 and within the DMN and reading comprehension (Buckner et al., 2008) and
below), 131 (37%) are in the borderline clinical range (65 − 69), and discourse (Aboud et al., 2016). It was also suggested that the deactiva­
111 (32%) are in the clinical range (70 and above) (Achenbach, 2009; tion of sub-parts of the DMN is related to different cognitive/linguistic
Achenbach and Ruffle, 2000). In addition to the CBCL questionnaire, the functions tightly linked to reading, such as visual attention (anterior
parents also completed the Kiddie Schedule for Affective Disorders and PCC), speech production (posterior ventral Medial Prefrontal Cortex),
Schizophrenia for DSM-5 (KSADS-5) diagnostic questionnaire (Kaufman perception and naming of objects (right inferior parietal cortex) (Seghier
et al., 2021), which includes a diagnostic criterion for ADHD. According and Price, 2012). It may be that this over-recruitment of the left hemi­
to the results, 180 (52%) children in this group were diagnosed with sphere, which is generally related to reading ability and changes in
ADHD. Interestingly, as a group, neurobiological patterns related to engagement of subregions within the DMN associated with semantic
ADHD were demonstrated. processing (Seghier and Price, 2012) in these groups, interferes with its
Following the higher comorbidity of this class, other scales from the role also in technical reading as was measured in the current study.
CBCL questionnaire were also tested (Table S12 and S13). The attention
problems scale also showed high comorbidity: 112 (32.46%) are in the 4.8. Big data in neuroimaging studies
normal range, 128 (36.99%) are in the clinical borderline clinical range,
and 106 (30.63%), as was expected. In all other scales, most of the class One of the caveats in neuroimaging studies is the difficulty repro­
was in the normal range, but unexpectedly more subjects were in the ducing study procedures and results (Botvinik-Nezer et al., 2020;
clinical range compared to other classes. Klapwijk et al., 2021), especially for brain-wide association studies
(Marek et al., 2022) as most studies are done using a limited number of
4.6. Do higher inhibition abilities characterize children with high socio- participants (Klapwijk et al., 2021; Marek et al., 2022). Decreased re­
economic status? sults variability can be achieved using a large sample size which di­
minishes the variability with a sample size greater than 2000
In the current study, we have created the four groups using the latent participants (Marek et al., 2020). In the current study, we used the data
profile analysis, which was based on the inhibition behavioral tasks. The from the Adolescent Brain Cognitive Development (ABCD)
data demonstrate how a higher number of females also characterizes (https://abcdstudy.org/) study, the largest children dataset aiming at
children in the “control” group (aka class2), a higher number of collecting various measures and assessments (neurocognition tasks,
Caucasian individuals, higher income, higher parental education, and neuroimaging scans, culture and environment questionnaires, physical
less motion inside the scanner compared to the other classes (see also and mental health, and biospecimens data) from pre-adolescents into
Table 3). An interesting question can be raised regarding the overall adulthood. One of the advantages of using large datasets like the ABCD
profile of children with high inhibition abilities- are these the charac­ is the ability to detect associations with many developmental outcomes
teristics of children from a high socio-economic status, i.e. affluent (Dick et al., 2020; Klapwijk et al., 2021), even if these large datasets are
population? On the one hand, this group’s high demographic profile usually associated with small effect sizes (Dick et al., 2020). Additional
may result in higher inhibition test scores (per(Moilanen et al., 2010)). studies using this or similar large datasets have the potential to fuel
On the other hand, even within the low SES group, children’s inhibition prediction models associated with emotional and cognitive outcomes for

13
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

children. Declaration of Competing Interest

5. Limitations The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
The results of the current study should be taken into consideration the work reported in this paper.
with the following limitations: first, minor variations of tests related to
inhibition were included in the analysis. Other tests that measure Data Availability
additional aspects of inhibitory control may reveal more fine-tuned
classes, especially in class1 (which includes the behavioral inhibition The data from this manuscript was downloaded from the ABCD
children), which comprised a large portion of the sample. Second, the dataset, which is publicly available. Scripts used for the analyses will be
recently developed algorithm for revealing the common correlation made available upon request.
matrix was validated and tested on a smaller sample (n = 458). It is
unknown to what extent the execution of the algorithm on large sam­ Acknowledgments
ples, as in our study, affects the results. Third, the interpretation of the
results was based solely on the differences between the control group Data used in the preparation of this article were obtained from the
and the other groups generated from the analysis, and therefore, the Adolescent Brain Cognitive Development (ABCD) Study (https://
actual values of the FC per group were not discussed. That is, there was abcdstudy.org), held in the NIMH Data Archive (NDA). This is a multi­
no difference if the values in the two compared classes were high (e.g. site, longitudinal study designed to recruit more than 10,000 children
close to 1) or low (close to zero). Lastly, to ensure that children in 2nd, age 9–10 and follow them over 10 years into early adulthood. The ABCD
6th and 7th grades (outliers) do not affect the cognitive profiles, these Study® is supported by the National Institutes of Health and additional
children were removed from the analysis, and the profiles were recre­ federal partners under award numbers U01DA041048, U01DA050989,
ated. The cognitive profiles remained similar to the original analysis as U01DA051016, U01DA041022, U01DA051018, U01DA051037,
well as the differences in reading measures between the groups. How­ U01DA050987, U01DA041174, U01DA041106, U01DA041117,
ever, when examining the FC differences within the DMN between the U01DA041028, U01DA041134, U01DA050988, U01DA051039,
four classes, different results from the original analysis were found (see U01DA041156, U01DA041025, U01DA041120, U01DA051038,
Supplemental material 1 in appendix B). One possibility for this differ­ U01DA041148, U01DA041093, U01DA041089, U24DA041123,
ence in DMN profiles is the multiple seeds ranging between lobes that U24DA041147. A full list of supporters is available at https://abcdstudy.
are part of the DMN (41 ROIs), which probably also contribute to the org/federal-partners.html. A listing of participating sites and a complete
multiple roles this network plays in numerous cognitive and emotional listing of the study investigators can be found at https://abcdstudy.org/
processes (Mak et al., 2017). As such, it may be that changes in the consortium_members/. ABCD consortium investigators designed and
number of participants within each group increase the variability within implemented the study and/or provided data but did not necessarily
the network and hence result in different profiles. Therefore, choosing a participate in the analysis or writing of this report. This manuscript
different network with a more restricted function (i.e. sensory network, reflects the views of the authors and may not reflect the opinions or
the cingulo-opercular or fronto-parietal ones) might have resulted in views of the NIH or ABCD consortium investigators.
more homogenous profiles even in the case of changing the number of The ABCD data repository grows and changes over time. The ABCD
participants. However, choosing these networks will disable the ability data used in this report came from ABCD collection 3165 and the Annual
to discuss the results in both emotional and cognitive contexts, as can be Release 2.0.1, DOI https://doi.org/https://doi.org/10.15154/1503209.
done when examining the DMN. Last but not least, failing to respond
correctly to the SST used in the current study is traditionally related to
inhibition challenges (Verbruggen and Logan, 2008) but also to chal­ Funding
lenges in attention and cognitive control (more broadly) (Matzke et al.,
2017). Therefore, the groups showing challenges in the SST may also This work was supported by the National Institute of Child Health
share an overall challenge in attention and cognitive control. and Human Development (NICHD) [grant numbers 5R01HD086011, PI:
Horowitz-Kraus].
6. Conclusions
Appendix A. Supporting information
This study demonstrated how in a community sample of children,
different subgroups could be found with inhibitory control difficulties Supplementary data associated with this article can be found in the
and how each of them had a different neurobiological signature within online version at doi:10.1016/j.dcn.2023.101198.
the DMN. These findings can contribute to the understanding of the
children’s variety of difficulties, which are not necessarily diagnosed References
with a particular learning disability, and the development of more
precise teaching methods. In addition, the findings can contribute to the Abo-elhija, D., Farah, R., Horowitz-Kraus, T., 2022. Stroop performance is related to
understanding of the neural basis of inhibitory control and aim to assist reading profiles in Hebrew-speaking individuals with dyslexia and typical readers.
Dyslexia 28 (2), 212–227.
in objective neuroimaging diagnosis in the future. As we examined only
Aboud, K.S., Bailey, S.K., Petrill, S.A., Cutting, L.E., 2016. Comprehending text versus
the DMN, additional studies are needed to examine the involvement of reading words in young readers with varying reading ability: distinct patterns of
other networks in inhibition, in addition to studies involving other as­ functional connectivity from common processing hubs. Dev. Sci. 19 (4), 632–656.
https://doi.org/10.1111/DESC.12422.
pects of inhibitory control. Finally, as the current study found different
Achenbach, T.M. (2009). The Achenbach system of empirically based assessment
profiles for children with inhibition abilities linked to reading skills (for (ASEBA): Development, findings, theory, and applications. University of Vermont,
example, see (Fuchs et al., 2020; Horowitz-Kraus et al., 2015a, 2015b; Research Center for Children, Youth, & Families.
Horowitz-Kraus and Holland, 2015; Peng and Goodrich, 2020), it will be Achenbach, T.M., Ruffle, T.M., 2000. The child behavior checklist and related forms for
assessing behavioral/emotional problems and competencies. Pediatr. Rev. 21 (8),
interesting to examine whether reading interventions with embedded 265–271.
elements of EF, especially inhibitory control, are helpful for reading Adlof, S.M., Catts, H.W., Little, T.D., 2006. Should the simple view of reading include a
improvement in particular profiles. fluency component? Read. Writ. 19 (9), 933–958.
Agcaoglu, O., Miller, R., Mayer, A.R., Hugdahl, K., Calhoun, V.D., 2015. Lateralization of
resting state networks and relationship to age and gender. NeuroImage 104,
310–325. https://doi.org/10.1016/j.neuroimage.2014.09.001.

14
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Allan, N.P., Hume, L.E., Allan, D.M., Farrington, A.L., Lonigan, C.J., 2014. Relations Diedenhofen, B., Musch, J., 2015. Cocor: a comprehensive solution for the statistical
between inhibitory control and the development of academic skills in preschool and comparison of correlations. PLoS ONE 10 (4), e0121945. https://doi.org/10.1371/
kindergarten: a meta-analysis. Dev. Psychol. 50 (10), 2368–2379. https://doi.org/ journal.pone.0121945.
10.1037/A0037493. Dosenbach, N.U.F., Fair, D.A., Cohen, A.L., Schlaggar, B.L., Petersen, S.E., 2008. A dual-
AnonABCD Research Consortium. (2019). Release Notes: Adolescent Brain Cognitive networks architecture of top-down control. Trends Cogn. Sci. 12 (3), 99–105.
Development (ABCD) Data Release 2.0. 1–29. https://doi.org/https://doi.org/ https://doi.org/10.1016/j.tics.2008.01.001.
10.15154/1503209. Dosenbach, N.U.F., Koller, J.M., Earl, E.A., Miranda-Dominguez, O., Klein, R.L., Van, A.
Aron, A.R., 2007. The neural basis of inhibition in cognitive control. In: In N., Snyder, A.Z., Nagel, B.J., Nigg, J.T., Nguyen, A.L., Wesevich, V., Greene, D.J.,
Neuroscientist, Vol. 13. Sage PublicationsSage CA: Los Angeles CA, pp. 214–228. Fair, D.A., 2017. Real-time motion analytics during brain MRI improve data quality
https://doi.org/10.1177/1073858407299288. and reduce costs. NeuroImage 161, 80. https://doi.org/10.1016/J.
Aron, A.R., Robbins, T.W., Poldrack, R.A., 2004. Inhibition and the right inferior frontal NEUROIMAGE.2017.08.025.
cortex. In: In Trends in Cognitive Sciences, Vol. 8. Trends Cogn Sci, pp. 170–177. Doucet, G.E., Janiri, D., Howard, R., O’Brien, M., Andrews-Hanna, J.R., Frangou, S.,
https://doi.org/10.1016/j.tics.2004.02.010. 2020. Transdiagnostic and disease-specific abnormalities in the default-mode
Bartholomew, M.E., Heller, W., Miller, G.A., 2021. Inhibitory control of emotional network hubs in psychiatric disorders: a meta-analysis of resting-state functional
processing: Theoretical and empirical considerations. In: In International Journal of imaging studies. Eur. Psychiatry 63 (1). https://doi.org/10.1192/j.eurpsy.2020.57.
Psychophysiology, Vol. 163. Elsevier B.V, pp. 5–10. https://doi.org/10.1016/j. Doyle, C., Smeaton, A.F., Roche, R.A.P., Boran, L., 2018. Inhibition and updating, but not
ijpsycho.2019.03.015. switching, predict developmental dyslexia and individual variation in reading
Binder, J.R., Medler, D.A., Desai, R., Conant, L.L., Liebenthal, E., 2005. Some ability. Front. Psychol. 9, 795. https://doi.org/10.3389/FPSYG.2018.00795/
neurophysiological constraints on models of word naming. NeuroImage 27 (3), BIBTEX.
677–693. https://doi.org/10.1016/J.NEUROIMAGE.2005.04.029. Duffy, K.A., Rosch, K.S., Nebel, M.B., Seymour, K.E., Lindquist, M.A., Pekar, J.J.,
Blair, C., Razza, R.P., 2007. Relating effortful control, executive function, and false belief Mostofsky, S.H., Cohen, J.R., 2021. Increased integration between default mode and
understanding to emerging math and literacy ability in kindergarten. Child Dev. 78 task-relevant networks in children with ADHD is associated with impaired response
(2), 647–663. control. Dev. Cogn. Neurosci. 50, 100980 https://doi.org/10.1016/J.
Borella, E., Carretti, B., Pelegrina, S., 2010. The specific role of inhibition in reading DCN.2021.100980.
comprehension in good and poor comprehenders. J. Learn. Disabil. 43 (6), 541–552. Eriksen, B.A., Eriksen, C.W., 1974. Effects of noise letters upon the identification of a
Botvinik-Nezer, R., Holzmeister, F., Camerer, C.F., Dreber, A., Huber, J., Johannesson, target letter in a nonsearch task. Percept. Psychophys. 16 (1), 143–149.
M., Kirchler, M., Iwanir, R., Mumford, J.A., & Adcock, R.A. (2020). Variability in the Esposito, R., Cieri, F., Chiacchiaretta, P., Cera, N., Lauriola, M., Di Giannantonio, M.,
analysis of a single neuroimaging dataset by many teamsBotvinik-Nezer, R., Tartaro, A., Ferretti, A., 2018. Modifications in resting state functional
Holzmeister, F., Camerer, C. F., Dreber, A., Huber, J., Johannesson, M., Kirchler, M., anticorrelation between default mode network and dorsal attention network:
Iwanir, R., Mumford, J. A., & Adcock, R. A. (2020). Variability in the ana. Nature, comparison among young adults, healthy elders and mild cognitive impairment
1–7. patients. Brain Imaging Behav. 12 (1), 127–141. https://doi.org/10.1007/s11682-
Buckner, R.L., 2012. The serendipitous discovery of the brain’s default network. 017-9686-y.
NeuroImage 62 (2), 1137–1145. https://doi.org/10.1016/j. Espy, K.A., McDiarmid, M.M., Cwik, M.F., Stalets, M.M., Hamby, A., Senn, T.E., 2004.
neuroimage.2011.10.035. The contribution of executive functions to emergent mathematic skills in preschool
Buckner, R.L., Andrews-Hanna, J.R., Schacter, D.L., 2008. The brain’s default network: children. Dev. Neuropsychol. 26 (1), 465–486.
anatomy, function, and relevance to disease. Ann. N. Y. Acad. Sci. 1124, 1–38. Fair, D.A., Cohen, A.L., Dosenbach, N.U.F., Church, J.A., Miezin, F.M., Barch, D.M.,
https://doi.org/10.1196/ANNALS.1440.011. Raichle, M.E., Petersen, S.E., Schlaggar, B.L., 2008. The maturing architecture of the
Butterfuss, R., Kendeou, P., 2018. The role of executive functions in reading brain’s default network. Proc. Natl. Acad. Sci. USA 105 (10), 4028–4032. https://
comprehension. Educ. Psychol. Rev. 30 (3), 801–826. doi.org/10.1073/pnas.0800376105.
Carver, C.S., White, T.L., 1994. Behavioral inhibition, behavioral activation, and Fan, F., Liao, X., Lei, T., Zhao, T., Xia, M., Men, W., Wang, Y., Hu, M., Liu, J., Qin, S.,
affective responses to impending reward and punishment: the BIS/BAS scales. Tan, S., Gao, J.H., Dong, Q., Tao, S., He, Y., 2021. Development of the default-mode
J. Personal. Soc. Psychol. 67 (2), 319. network during childhood and adolescence: A longitudinal resting-state fMRI study.
Casaletto, K.B., Umlauf, A., Beaumont, J., Gershon, R., Slotkin, J., Akshoomoff, N., NeuroImage 226, 117581. https://doi.org/10.1016/j.neuroimage.2020.117581.
Heaton, R.K., 2015. Demographically corrected normative standards for the English Fan, J., Gan, J., Liu, W., Zhong, M., Liao, H., Zhang, H., Yi, J., Chan, R.C.K., Tan, C.,
version of the NIH Toolbox Cognition Battery. J. Int. Neuropsychol. Soc. 21 (5), Zhu, X., 2018. Resting-state default mode network related functional connectivity is
378–391. associated with sustained attention deficits in schizophrenia and obsessive-
Casey, B.J., Cannonier, T., Conley, M.I., Cohen, A.O., Barch, D.M., Heitzeg, M.M., compulsive disorder. Front. Behav. Neurosci. 12. https://doi.org/10.3389/
Soules, M.E., Teslovich, T., Dellarco, D.V., Garavan, H., Orr, C.A., Wager, T.D., fnbeh.2018.00319.
Banich, M.T., Speer, N.K., Sutherland, M.T., Riedel, M.C., Dick, A.S., Bjork, J.M., Feczko, E., Earl, E., Perrone, A., Fair, D. (2020). ABCD Reproducible Matched Samples
Thomas, K.M., Workgroup, A.I.A., 2018. The Adolescent Brain Cognitive (ARMS) software. osf.io/7xn4f.
Development (ABCD) study: Imaging acquisition across 21 sites. Dev. Cogn. Fox, N.A., Henderson, H.A., Marshall, P.J., Nichols, K.E., Ghera, M.M., 2005. Behavioral
Neurosci. 32, 43–54. https://doi.org/10.1016/j.dcn.2018.03.001. inhibition: Linking biology and behavior within a developmental framework. Annu.
Castellanos, F.X., Proal, E., 2012. Large-scale brain systems in ADHD: beyond the Rev. Psychol. 56, 235–262.
prefrontal–striatal model. Trends Cogn. Sci. 16 (1), 17–26. Fryer, S.L., Roach, B.J., Ford, J.M., Donaldson, K.R., Calhoun, V.D., Pearlson, G.D.,
Chall, J.S., 1983. Stages of reading development. McGraw-Hill,. Kiehl, K.A., Srihari, V.H., McGlashan, T.H., Woods, S.W., Mathalon, D.H., 2018.
Clark, C.A.C., Pritchard, V.E., Woodward, L.J., 2010. Preschool executive functioning Should I stay or should I go? FMRI study of response inhibition in early illness
abilities predict early mathematics achievement. Dev. Psychol. 46 (5), 1176. Schizophrenia and risk for psychosis. Schizophr. Bull. 45 (1), 158–168. https://doi.
Cohen, J., 1977. Differences between Correlation Coefficients. Statistical Power Analysis org/10.1093/schbul/sbx198.
for the Behavioral Sciences. Elsevier, pp. 109–143. https://doi.org/10.1016/c2013- Fuchs, L., Fuchs, D., Seethaler, P.M., Barnes, M.A., 2020. Addressing the role of working
0-10517-x. memory in mathematical word-problem solving when designing intervention for
Conners, F.A., 2009. Attentional control and the simple view of reading. Read. Writ. 22 struggling learners. ZDM 52 (1), 87–96.
(5), 591–613. Gershon, R.C., Slotkin, J., Manly, J.J., Blitz, D.L., Beaumont, J.L., Schnipke, D., Wallner-
Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven Allen, K., Golinkoff, R.M., Gleason, J.B., Hirsh-Pasek, K., 2013. IV. NIH Toolbox
attention in the brain. Nat. Rev. Neurosci. 3 (3), 201–215. https://doi.org/10.1038/ Cognition Battery (CB): Measuring language (vocabulary comprehension and
nrn755. reading decoding). Monogr. Soc. Res. Child Dev. 78 (4), 49–69.
Corbetta, M., Patel, G., Shulman, G.L., 2008. The Reorienting System of the Human Gordon, E.M., Laumann, T.O., Adeyemo, B., Huckins, J.F., Kelley, W.M., Petersen, S.E.,
Brain: From Environment to Theory of Mind. In: In Neuron, Vol. 58. NIH Public 2016. Generation and evaluation of a cortical area Parcellation from resting-state
Access, pp. 306–324. https://doi.org/10.1016/j.neuron.2008.04.017. correlations. Cereb. Cortex 26 (1), 288–303. https://doi.org/10.1093/cercor/
Corder, G.W., Foreman, D. l, 2009. Nonparametric statistics: A step-by-step approach. bhu239.
John Wiley & Sons. Gough, P.B., Tunmer, W.E., 1986. Decoding, reading, and reading disability. Remedial
Cragg, L., Gilmore, C., 2014. Skills underlying mathematics: The role of executive Spec. Educ. 7, 6–10.
function in the development of mathematics proficiency. Trends Neurosci. Educ. 3 Haft, S., Caballero, J.N., Tanak, H., Zekelman, N., Cutting, L.E., Uchiko, Y., Hoeft, F.,
(2), 63–68. https://doi.org/10.1016/J.TINE.2013.12.001. 2019. Direct and indirect contributions of executive function to word decoding and
Dajani, D.R., Llabre, M.M., Nebel, M.B., Mostofsky, S.H., Uddin, L.Q., 2016. reading comprehension in kindergarten. Learn. Individ. Differ. 76, 101783.
Heterogeneity of executive functions among comorbid neurodevelopmental Hagler Jr., D.J., Hatton, S., Cornejo, M.D., Makowski, C., Fair, D.A., Dick, A.S.,
disorders. Sci. Rep. 6, 36566. Sutherland, M.T., Casey, B.J., Barch, D.M., Harms, M.P., 2019. Image processing and
Dajani, D.R., Burrows, C.A., Odriozola, P., Baez, A., Nebel, M.B., Mostofsky, S.H., analysis methods for the Adolescent Brain Cognitive Development Study.
Uddin, L.Q., 2019. Investigating functional brain network integrity using a Neuroimage 202, 116091.
traditional and novel categorical scheme for neurodevelopmental disorders. Hanakawa, T., Immisch, I., Toma, K., Dimyan, M.A., Van Gelderen, P., Hallett, M., 2003.
NeuroImage: Clin. 21. https://doi.org/10.1016/j.nicl.2019.101678. Functional properties of brain areas associated with motor execution and imagery.
Diamond, A., 2013. Executive functions. Annu. Rev. Psychol. 64, 135–168. J. Neurophysiol. 89 (2), 989–1002. https://doi.org/10.1152/jn.00132.2002.
Dick, A.S., Watts, A.L., Heeringa, S., Lopez, D.A., Bartsch, H., Fan, C.C., Palmer, C., Harikumar, A., Evans, D.W., Dougherty, C.C., Carpenter, K.L.H., Michael, A.M., 2021.
Reuter, C., Marshall, A., Haist, F., Hawes, S., Nichols, T.E., Barch, D.M., Jernigan, T. A review of the default mode network in autism spectrum disorders and attention
L., Garavan, H., Grant, S., Pariyadath, V., Hoffman, E., Neale, M., Thompson, W.K., deficit hyperactivity disorder. Brain Connect. 11 (4), 253–263. https://doi.org/
2020. Meaningful effects in the adolescent brain cognitive development study. 10.1089/brain.2020.0865.
BioRxiv. https://doi.org/10.1101/2020.09.01.276451.

15
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Harnishfeger, K.K., 1995. The development of cognitive inhibition: theories, definitions, neurocognition battery. Dev. Cogn. Neurosci. 32, 67–79. https://doi.org/10.1016/j.
and research evidence. Interf. Inhib. Cogn. 175–204. https://doi.org/10.1016/B978- dcn.2018.02.006.
012208930-5/50007-6. Mak, L.E., Minuzzi, L., MacQueen, G., Hall, G., Kennedy, S.H., Milev, R., 2017. The
Harnishfeger, K.K., Bjorklund, D.F., 1994. A developmental perspective on individual default mode network in healthy individuals: a systematic review and meta-analysis.
differences in inhibition. Learn. Individ. Differ. 6 (3), 331–355. https://doi.org/ Brain Connect. 7 (1), 25–33.
10.1016/1041-6080(94)90021-3. Malouin, F., Richards, C.L., Jackson, P.L., Dumas, F., Doyon, J., 2003. Brain activations
Hernández-Álvarez, D.M., Pacheco, L., Velasco-Segura, R., Pérez de la Mora, M., Tejeda- during motor imagery of locomotor-related tasks: A PET study. Hum. Brain Mapp. 19
Romero, C., González-García, N., 2020. Default mode network efficiency is (1), 47–62. https://doi.org/10.1002/hbm.10103.
correlated with deficits in inhibition in adolescents with inhalant use disorder. Front. Marek, S., Tervo-Clemmens, B., Calabro, F.J., Montez, D.F., Kay, B.P., Hatoum, A.S.,
Psychiatry 11, 209. Donohue, M.R., Foran, W., Miller, R.L., & Feczko, E. (2020). Towards Reproducible
Horowitz-Kraus, T., Holland, S.K., 2015. Greater functional connectivity between Brain-Wide Association Studies. BioRxiv.
reading and error-detection regions following training with the reading acceleration Marek, S., Tervo-Clemmens, B., Calabro, F.J., Montez, D.F., Kay, B.P., Hatoum, A.S.,
program in children with reading difficulties. Ann. Dyslexia 65 (1), 1–23. Donohue, M.R., Foran, W., Miller, R.L., Hendrickson, T.J., 2022. Reproducible brain-
Horowitz-Kraus, T., Toro-Serey, C., DiFrancesco, M., 2015b. Increased resting-state wide association studies require thousands of individuals. Nature 603 (7902),
functional connectivity in the cingulo-opercular cognitive-control network after 654–660.
intervention in children with reading difficulties. PloS One 10 (7), e0133762. Margulies, D.S., Vincent, J.L., Kelly, C., Lohmann, G., Uddin, L.Q., Biswal, B.B.,
Horowitz-Kraus, T., DiFrancesco, M., Kay, B., Wang, Y., Holland, S.K., 2015a. Increased Villringer, A., Castellanos, F.X., Milham, M.P., Petrides, M., 2009. Precuneus shares
functional connectivity of specific brain networks after reading training in dyslexic intrinsic functional architecture in humans and monkeys. Proc. Natl. Acad. Sci. USA
children. Clin. Neuroimage 8, 619–630. 106 (47), 20069–20074. https://doi.org/10.1073/pnas.0905314106.
Horowitz-Kraus, T., Farah, R., Hajinazarian, A., Eaton, K., Rajagopal, A., Schmithorst, V. Matzke, D., Hughes, M., Badcock, J.C., Michie, P., Heathcote, A., 2017. Failures of
J., 2017. Maturation of brain regions related to the default mode network during cognitive control or attention? The case of stop-signal deficits in schizophrenia.
adolescence facilitates narrative comprehension. J. Child Adolesc. Behav. 5, Atten., Percept., Psychophys. 79 (4), 1078–1086. https://doi.org/10.3758/s13414-
328–336. 017-1287-8.
Husain, M., Nachev, P., 2007. Space and the parietal cortex. In: In Trends in Cognitive Meixner, J.M., Warner, G.J., Lensing, N., Schiefele, U., Elsner, B., 2019. The relation
Sciences, Vol. 11. Elsevier Current Trends., pp. 30–36. https://doi.org/10.1016/j. between executive functions and reading comprehension in primary-school students:
tics.2006.10.011 A cross-lagged-panel analysis. Early Child. Res. Q. 46, 62–74. https://doi.org/
Johann, V., Könen, T., Karbach, J., 2020. The unique contribution of working memory, 10.1016/J.ECRESQ.2018.04.010.
inhibition, cognitive flexibility, and intelligence to reading comprehension and Moilanen, K.L., Shaw, D.S., Dishion, T.J., Gardner, F., Wilson, M., 2010. Predictors of
reading speed. Child Neuropsychol. 26 (3), 324–344. longitudinal growth in inhibitory control in early childhood. Soc. Dev. 19 (2),
Karcher, N.R., O’Brien, K.J., Kandala, S., Barch, D.M., 2019. Resting State Functional 326–347.
Connectivity and Psychotic-Like Experiences in Childhood: Results from the Muthén, L.K.and Muthén, B.O. (1998–2017). Mplus User’s Guide. Eighth Edition.
Adolescent Brain Cognitive Development Study. Biol. Psychiatry 86 (1), 7. https:// Neta, M., Schlaggar, B., Petersen, S., 2014. Separable responses to error. Ambiguity,
doi.org/10.1016/J.BIOPSYCH.2019.01.013. React. Time Cingulo-Opercular Task. Control Reg. Vol. 99 https://doi.org/10.1016/
Kassambara, A. (2020). rstatix: Pipe-Friendly Framework for Basic Statistical Tests. (R j.neuroimage.2014.05.053.
package version 0.6.0.). https://cran.r-project.org/package=rstatix. Ng, K.K., Lo, J.C., Lim, J.K.W., Chee, M.W.L., Zhou, J., 2016. Reduced functional
Kaufman, J., Kobak, K., Birmaher, B., de Lacy, N., 2021. KSADS-COMP Perspectives on segregation between the default mode network and the executive control network in
Child Psychiatric Diagnostic Assessment and Treatment Planning, 60. Journal of the healthy older adults: a longitudinal study. NeuroImage 133, 321–330. https://doi.
American Academy of Child & Adolescent Psychiatry, pp. 540–542. https://doi.org/ org/10.1016/j.neuroimage.2016.03.029.
10.1016/J.JAAC.2020.08.470. Nielsen, J.A., Zielinski, B.A., Ferguson, M.A., Lainhart, J.E., Anderson, J.S., 2013. An
Khalilzadeh, J., Tasci, A.D.A., 2017. Large sample size, significance level, and the effect evaluation of the left-brain vs. right-brain hypothesis with resting state functional
size: Solutions to perils of using big data for academic research. Tour. Manag. 62, connectivity magnetic resonance imaging. PLoS ONE 8 (8), 71275. https://doi.org/
89–96. https://doi.org/10.1016/j.tourman.2017.03.026. 10.1371/journal.pone.0071275.
Kim, Y.-S.G., 2020. Simple but not simplistic: the simple view of reading unpacked and Nylund, K.L., & Muthén, B.O. (2007). Deciding on the Number of Classes in Latent Class
expanded. Read. Leag. J. 1 (2), 15–34. Analysis and Growth Mixture Modeling: A Monte Carlo Simulation Study. 14(4),
Klapwijk, E.T., van den Bos, W., Tamnes, C.K., Raschle, N.M., Mills, K.L., 2021. 535–569.
Opportunities for increased reproducibility and replicability of developmental Owens, M.M., Yuan, D., Hahn, S., Albaugh, M., Allgaier, N., Chaarani, B., Potter, A.,
neuroimaging. Dev. Cogn. Neurosci. 47 (2020), 100902 https://doi.org/10.1016/j. Garavan, H., 2020. Investigation of psychiatric and neuropsychological correlates of
dcn.2020.100902. default mode network and dorsal attention network anticorrelation in children.
Klenberg, L., Korkman, M., Lahti-Nuuttila, P., 2001. Differential development of Cereb. Cortex 30 (12), 6083–6096. https://doi.org/10.1093/cercor/bhaa143.
attention and executive functions in 3- to 12-year-old finnish children. Dev. Pacheco, D., Owen, M., Caughy, M., 2018. Growth in inhibitory control among low-
Neuropsychol. 20 (1), 407–428. https://doi.org/10.1207/S15326942DN2001_6. income, ethnic-minority preschoolers: a group-based modeling approach. Early
Kolb, B., Gibb, R., 2011. Brain plasticity and behaviour in the developing brain. J. Can. Child. Res. Q. 42, 247–255. https://doi.org/10.1016/j.ecresq.2017.10.010.
Acad. Child Adolesc. Psychiatry 20 (4), 265. /pmc/articles/PMC3222570/. Pastor, D.A., Barron, K.E., Miller, B.J., Davis, S.L., 2007. A latent profile analysis of
Koziol, L.F., Budding, D.E., Chidekel, D., 2013. ADHD as a Model of Brain-behavior college students’ achievement goal orientation. Contemp. Educ. Psychol. 32 (1),
Relationships. Springer. 8–47.
LaBerge, D., Samuels, S.J., 1974. Toward a theory of automatic information processing in Peng, P., Goodrich, J.M., 2020. The cognitive element model of reading instruction.
reading. Cogn. Psychol. 6, 293–323. Read. Res. Q. 55, S77–S88.
Lanzoni, L., Ravasio, D., Thompson, H., Vatansever, D., Margulies, D., Smallwood, J., Power, J.D., Mitra, A., Laumann, T.O., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E., 2014.
Jefferies, E., 2020. The role of default mode network in semantic cue integration. Methods to detect, characterize, and remove motion artifact in resting state fMRI.
NeuroImage 219, 117019. https://doi.org/10.1016/J.NEUROIMAGE.2020.117019. NeuroImage 84, 320–341. https://doi.org/10.1016/J.NEUROIMAGE.2013.08.048.
Ledoit, O., & Wolf, M. (1996). A Well-Conditioned Estimator For Large Dimensional Ptak, R. (2012). The Frontoparietal Attention Network of the Human Brain (Vol. 18).
Covariance Matrices (Vol. 88). https://doi.org/10.1016/S0047–259X(03)00096–4. https://doi.org/10.1177/1073858411409051.
Leech, R., Kamourieh, S., Beckmann, C.F., Sharp, D.J., 2011. Fractionating the default Rohr, C.S., Arora, A., Cho, I.Y.K., Katlariwala, P., Dimond, D., Dewey, D., Bray, S., 2018.
mode network: distinct contributions of the ventral and dorsal posterior cingulate Functional network integration and attention skills in young children. Dev. Cogn.
cortex to cognitive control. J. Neurosci. 31 (9), 3217–3224. https://doi.org/ Neurosci. 30, 200–211. https://doi.org/10.1016/j.dcn.2018.03.007.
10.1523/JNEUROSCI.5626-10.2011. Rosenberg, J.M., Beymer, P.N., Anderson, D.J., Lissa, V., & Schmidt, J.A. (2018).
Lenhard, W.and Lenhard, A. (2014). Hypothesis tests for comparing correlations. tidyLPA: An R Package to Easily Carry Out Latent Profile Analysis ( LPA) Using
Bibergau, Germany: Psychometrica. Open-Source or Commercial Software A tidy user-interface. 3, 3–5. https://doi.org/
Li, C.S.R., Yan, P., Bergquist, K.L., Sinha, R., 2007. Greater activation of the “default” 10.21105/joss.00978.
brain regions predicts stop signal errors. NeuroImage 38 (3), 640–648. https://doi. Sarsour, K., Sheridan, M., Jutte, D., Nuru-Jeter, A., Hinshaw, S., Boyce, W.T., 2011.
org/10.1016/j.neuroimage.2007.07.021. Family socioeconomic status and child executive functions: the roles of language,
Li, W., Mai, X., Liu, C., 2014. The default mode network and social understanding of home environment, and single parenthood. J. Int. Neuropsychol. Soc. 17 (1),
others: what do brain connectivity studies tell us. Front. Hum. Neurosci. 8 (1 FEB), 120–132. https://doi.org/10.1017/S1355617710001335.
74. https://doi.org/10.3389/fnhum.2014.00074. Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Kenna, H., Reiss, A.L.,
Liddle, E.B., Hollis, C., Batty, M.J., Groom, M.J., Totman, J.J., Liotti, M., Scerif, G., Greicius, M.D., 2007. Dissociable intrinsic connectivity networks for salience
Liddle, P.F., 2011. Task-related default mode network modulation and inhibitory processing and executive control. J. Neurosci. 27 (9), 2349–2356. https://doi.org/
control in ADHD: Effects of motivation and methylphenidate. J. Child Psychol. 10.1523/jneurosci.5587-06.2007.
Psychiatry 52 (7), 761–771. Seghier, M.L., Price, C.J., 2012. Functional heterogeneity within the default network
Logan, G.D. (1994). On the ability to inhibit thought and action: A users’ guide to the during semantic processing and speech production. Front. Psychol. 3 (AUG) https://
stop signal paradigm. doi.org/10.3389/FPSYG.2012.00281.
Lu, T., Pan, Y., Kao, S.-Y., Li, C., Kohane, I., Chan, J., Yankner, B.A., 2004. Gene Sidlauskaite, J., Sonuga-Barke, E., Roeyers, H., Wiersema, J.R., 2016. Altered intrinsic
regulation and DNA damage in the ageing human brain. Nat. 2004 429:6994 429 organisation of brain networks implicated in attentional processes in adult attention-
(6994), 883–891. https://doi.org/10.1038/nature02661. deficit/hyperactivity disorder: a resting-state study of attention, default mode and
Luciana, M., Bjork, J.M., Nagel, B.J., Barch, D.M., Gonzalez, R., Nixon, S.J., Banich, M.T., salience network connectivity. Eur. Arch. Psychiatry Clin. Neurosci. 266 (4),
2018. Adolescent neurocognitive development and impacts of substance use: 349–357.
Overview of the adolescent brain cognitive development (ABCD) baseline

16
V.K. Fekson et al. Developmental Cognitive Neuroscience 60 (2023) 101198

Singh-Curry, V., Husain, M., 2009. The functional role of the inferior parietal lobe in the Uddin, L.Q., Kelly, A.M.C., Biswal, B.B., Castellanos, F.X., Milham, M.P., 2009.
dorsal and ventral stream dichotomy. Neuropsychologia 47 (6), 1434–1448. https:// Functional connectivity of default mode network components: correlation,
doi.org/10.1016/j.neuropsychologia.2008.11.033. anticorrelation, and causality. Hum. Brain Mapp. 30 (2), 625–637. https://doi.org/
Spencer, M., Cutting, L.E., 2021. Relations among executive function, decoding, and 10.1002/hbm.20531.
reading comprehension: an investigation of sex differences. Discourse Process. 58 Verbruggen, F., Logan, G.D., 2008. Response inhibition in the stop-signal paradigm.
(1), 42–59. Trends Cogn. Sci. 12 (11), 418–424. https://doi.org/10.1016/j.tics.2008.07.005.
Spreng, R.N., Stevens, W.D., Chamberlain, J.P., Gilmore, A.W., Schacter, D.L., 2010. Vossel, S., Geng, J.J., Fink, G.R., 2014. Dorsal and ventral attention systems: distinct
Default network activity, coupled with the frontoparietal control network, supports neural circuits but collaborative roles. Neuroscientist 20 (2), 150–159. https://doi.
goal-directed cognition. NeuroImage 53 (1), 303–317. https://doi.org/10.1016/j. org/10.1177/1073858413494269.
neuroimage.2010.06.016. Welsh, M.C., Pennington, B.F., Groisser, D.B., 1991. A normative-developmental study of
Sripada, C., Kessler, D., Fang, Y., Welsh, R.C., Prem Kumar, K., Angstadt, M., 2014a. executive function: a window on prefrontal function in children. Dev. Neuropsychol.
Disrupted network architecture of the resting brain in attention-deficit/hyperactivity 7 (2), 131–149. https://doi.org/10.1080/87565649109540483.
disorder. Hum. Brain Mapp. 35 (9), 4693–4705. https://doi.org/10.1002/ Williams, J., Omizzolo, C., Galea, M.P., Vance, A., 2013. Motor imagery skills of children
hbm.22504. with attention deficit hyperactivity disorder and developmental coordination
Sripada, C., Kessler, D., Fang, Y., Welsh, R.C., Prem Kumar, K., Angstadt, M., 2014b. disorder. Hum. Mov. Sci. 32 (1), 121–135. https://doi.org/10.1016/j.
Disrupted network architecture of the resting brain in attention-deficit/hyperactivity humov.2012.08.003.
disorder. Hum. Brain Mapp. 35 (9), 4693–4705. Williams, G.A., & Kibowski, F. (2016). Latent class analysis and latent profile analysis.
Staffaroni, A.M., Brown, J.A., Casaletto, K.B., Elahi, F.M., Deng, J., Neuhaus, J., Handbook of Methodological Approaches to Community-Based Research:
Cobigo, Y., Mumford, P.S., Walters, S., Saloner, R., 2018. The longitudinal trajectory Qualitative, Quantitative, and Mixed Methods, 143–151.
of default mode network connectivity in healthy older adults varies as a function of Wirth, M., Jann, K., Dierks, T., Federspiel, A., Wiest, R., Horn, H., 2011. Semantic
age and is associated with changes in episodic memory and processing speed. memory involvement in the default mode network: a functional neuroimaging study
J. Neurosci. 38 (11), 2809–2817. using independent component analysis. NeuroImage 54 (4), 3057–3066. https://doi.
Supekar, K., Uddin, L.Q., Prater, K., Amin, H., Greicius, M.D., Menon, V., 2010. org/10.1016/J.NEUROIMAGE.2010.10.039.
Development of functional and structural connectivity within the default mode Wurpts, I.C., Geiser, C., 2014. Is adding more indicators to a latent class analysis
network in young children. NeuroImage 52 (1), 290–301. https://doi.org/10.1016/j. beneficial or detrimental? Results of a Monte-Carlo study. Front. Psychol. 5, 920.
neuroimage.2010.04.009. Zhang, H., Long, Z., Ge, R., Xu, L., Jin, Z., Yao, L., Liu, Y., 2014. Motor imagery learning
Swanson, N., Eichele, T., Pearlson, G., Kiehl, K., Yu, Q., Calhoun, V.D., 2011. Lateral modulates functional connectivity of multiple brain systems in resting state. PLoS
differences in the default mode network in healthy controls and patients with ONE 9 (1), e85489. https://doi.org/10.1371/journal.pone.0085489.
schizophrenia. Hum. Brain Mapp. 32 (4), 654–664. https://doi.org/10.1002/ Zhang, S., Li, C.S., 2012. Functional connectivity mapping of the human precuneus by
hbm.21055. resting state fMRI. NeuroImage 59 (4), 3548–3562. https://doi.org/10.1016/j.
Taboada Barber, A., Cartwright, K.B., Hancock, G.R., Klauda, S.L., 2021. Beyond the neuroimage.2011.11.023.
simple view of reading: The role of executive functions in emergent bilinguals’ and Zhitnikov, A., Mulayoff, R., Michaeli, T. (2018). Revealing common statistical behaviors
English monolinguals’ reading comprehension. Read. Res. Q. 56, S45–S64. in heterogeneous populations. 35th International Conference on Machine Learning,
Tein, J.Y., Coxe, S., Cham, H., 2013. Statistical power to detect the correct number of ICML 2018, 13, 9490–9499.
classes in latent profile analysis. Struct. Equ. Model. 20 (4), 640–657. https://doi.
org/10.1080/10705511.2013.824781.

17